public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug libstdc++/24071] solaris vs. __gthread_active_p
[not found] <bug-24071-365@http.gcc.gnu.org/bugzilla/>
@ 2005-10-02 21:21 ` mark at codesourcery dot com
2005-10-02 21:28 ` ebotcazou at gcc dot gnu dot org
` (8 subsequent siblings)
9 siblings, 0 replies; 10+ messages in thread
From: mark at codesourcery dot com @ 2005-10-02 21:21 UTC (permalink / raw)
To: gcc-bugs
------- Comment #8 from mark at codesourcery dot com 2005-10-02 21:21 -------
Subject: Re: solaris vs. __gthread_active_p
bkoz at gcc dot gnu dot org wrote:
> However, there is a bigger issue involved, which is still in progress. Actually,
> I'd like to broaden this out a bit since it looks like we are thinking of
> solving a related issue on linux as well. (See RH bugzilla, link below)
> https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=165728
>
> Therefore, I'm suggesting I close this, and open a more general enhancement bug.
Sure, that sounds fine.
My opinion is that, even if it means a small performance hit, we should
move the weak declarations of pthread_* out of headers that can be
included in user code. I understand that would be an ABI change, but
you're working on a new libstdc++ ABI, so you could do it there.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24071
^ permalink raw reply [flat|nested] 10+ messages in thread
* [Bug libstdc++/24071] solaris vs. __gthread_active_p
[not found] <bug-24071-365@http.gcc.gnu.org/bugzilla/>
2005-10-02 21:21 ` [Bug libstdc++/24071] solaris vs. __gthread_active_p mark at codesourcery dot com
@ 2005-10-02 21:28 ` ebotcazou at gcc dot gnu dot org
2005-10-09 17:57 ` [Bug target/24071] __gthread_active_p vs __gthread_once ebotcazou at gcc dot gnu dot org
` (7 subsequent siblings)
9 siblings, 0 replies; 10+ messages in thread
From: ebotcazou at gcc dot gnu dot org @ 2005-10-02 21:28 UTC (permalink / raw)
To: gcc-bugs
------- Comment #9 from ebotcazou at gcc dot gnu dot org 2005-10-02 21:27 -------
> Therefore, I'm suggesting I close this, and open a more general enhancement bug.
I'd rather keep it open and recategorize it as a "target" PR, since there is
really a problem with __gthread_active_p on Solaris because of the
Solaris/POSIX threads duality.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24071
^ permalink raw reply [flat|nested] 10+ messages in thread
* [Bug target/24071] __gthread_active_p vs __gthread_once
[not found] <bug-24071-365@http.gcc.gnu.org/bugzilla/>
2005-10-02 21:21 ` [Bug libstdc++/24071] solaris vs. __gthread_active_p mark at codesourcery dot com
2005-10-02 21:28 ` ebotcazou at gcc dot gnu dot org
@ 2005-10-09 17:57 ` ebotcazou at gcc dot gnu dot org
2005-12-14 11:47 ` neil at fnxweb dot com
` (6 subsequent siblings)
9 siblings, 0 replies; 10+ messages in thread
From: ebotcazou at gcc dot gnu dot org @ 2005-10-09 17:57 UTC (permalink / raw)
To: gcc-bugs
--
ebotcazou at gcc dot gnu dot org changed:
What |Removed |Added
----------------------------------------------------------------------------
AssignedTo|bkoz at gcc dot gnu dot org |unassigned at gcc dot gnu
| |dot org
Status|ASSIGNED |NEW
Component|libstdc++ |target
GCC build triplet|sparc-sun-solaris2.9 |*-*-solaris2.*
GCC host triplet|sparc-sun-solaris2.9 |*-*-solaris2.*
GCC target triplet|sparc-sun-solaris2.9 |*-*-solaris2.*
Summary|solaris vs. |__gthread_active_p vs
|__gthread_active_p |__gthread_once
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24071
^ permalink raw reply [flat|nested] 10+ messages in thread
* [Bug target/24071] __gthread_active_p vs __gthread_once
[not found] <bug-24071-365@http.gcc.gnu.org/bugzilla/>
` (2 preceding siblings ...)
2005-10-09 17:57 ` [Bug target/24071] __gthread_active_p vs __gthread_once ebotcazou at gcc dot gnu dot org
@ 2005-12-14 11:47 ` neil at fnxweb dot com
2006-10-28 5:36 ` ebotcazou at gcc dot gnu dot org
` (5 subsequent siblings)
9 siblings, 0 replies; 10+ messages in thread
From: neil at fnxweb dot com @ 2005-12-14 11:47 UTC (permalink / raw)
To: gcc-bugs
------- Comment #10 from neil at fnxweb dot com 2005-12-14 11:47 -------
For ref., I've just raised PR 25409 which may possible be a dup. of this
problem. It's nothing to do with Solaris, though, so I didn't just add the
details here.
--
neil at fnxweb dot com changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |neil at fnxweb dot com
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24071
^ permalink raw reply [flat|nested] 10+ messages in thread
* [Bug target/24071] __gthread_active_p vs __gthread_once
[not found] <bug-24071-365@http.gcc.gnu.org/bugzilla/>
` (3 preceding siblings ...)
2005-12-14 11:47 ` neil at fnxweb dot com
@ 2006-10-28 5:36 ` ebotcazou at gcc dot gnu dot org
2006-10-31 9:34 ` bkoz at gcc dot gnu dot org
` (4 subsequent siblings)
9 siblings, 0 replies; 10+ messages in thread
From: ebotcazou at gcc dot gnu dot org @ 2006-10-28 5:36 UTC (permalink / raw)
To: gcc-bugs
------- Comment #11 from ebotcazou at gcc dot gnu dot org 2006-10-28 05:36 -------
About to submit a patch.
--
ebotcazou at gcc dot gnu dot org changed:
What |Removed |Added
----------------------------------------------------------------------------
CC|ebotcazou at libertysurf dot|
|fr |
AssignedTo|unassigned at gcc dot gnu |ebotcazou at gcc dot gnu dot
|dot org |org
Status|NEW |ASSIGNED
Last reconfirmed|2005-10-01 18:45:54 |2006-10-28 05:36:21
date| |
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24071
^ permalink raw reply [flat|nested] 10+ messages in thread
* [Bug target/24071] __gthread_active_p vs __gthread_once
[not found] <bug-24071-365@http.gcc.gnu.org/bugzilla/>
` (4 preceding siblings ...)
2006-10-28 5:36 ` ebotcazou at gcc dot gnu dot org
@ 2006-10-31 9:34 ` bkoz at gcc dot gnu dot org
2006-10-31 17:55 ` ebotcazou at gcc dot gnu dot org
` (3 subsequent siblings)
9 siblings, 0 replies; 10+ messages in thread
From: bkoz at gcc dot gnu dot org @ 2006-10-31 9:34 UTC (permalink / raw)
To: gcc-bugs
------- Comment #12 from bkoz at gcc dot gnu dot org 2006-10-31 09:34 -------
Thanks for fixing this "for real" Eric.
-benjamin
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24071
^ permalink raw reply [flat|nested] 10+ messages in thread
* [Bug target/24071] __gthread_active_p vs __gthread_once
[not found] <bug-24071-365@http.gcc.gnu.org/bugzilla/>
` (5 preceding siblings ...)
2006-10-31 9:34 ` bkoz at gcc dot gnu dot org
@ 2006-10-31 17:55 ` ebotcazou at gcc dot gnu dot org
2006-10-31 17:56 ` ebotcazou at gcc dot gnu dot org
` (2 subsequent siblings)
9 siblings, 0 replies; 10+ messages in thread
From: ebotcazou at gcc dot gnu dot org @ 2006-10-31 17:55 UTC (permalink / raw)
To: gcc-bugs
------- Comment #13 from ebotcazou at gcc dot gnu dot org 2006-10-31 17:55 -------
Subject: Bug 24071
Author: ebotcazou
Date: Tue Oct 31 17:54:58 2006
New Revision: 118259
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=118259
Log:
PR target/24071
* gthr-posix.h (__gthread_active_p): New implementation on Solaris.
* gthr-posix95.h (__gthread_active_p): Likewise.
Modified:
trunk/gcc/ChangeLog
trunk/gcc/gthr-posix.h
trunk/gcc/gthr-posix95.h
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24071
^ permalink raw reply [flat|nested] 10+ messages in thread
* [Bug target/24071] __gthread_active_p vs __gthread_once
[not found] <bug-24071-365@http.gcc.gnu.org/bugzilla/>
` (6 preceding siblings ...)
2006-10-31 17:55 ` ebotcazou at gcc dot gnu dot org
@ 2006-10-31 17:56 ` ebotcazou at gcc dot gnu dot org
2006-10-31 18:03 ` ebotcazou at gcc dot gnu dot org
2006-11-30 1:23 ` chaoyingfu at gcc dot gnu dot org
9 siblings, 0 replies; 10+ messages in thread
From: ebotcazou at gcc dot gnu dot org @ 2006-10-31 17:56 UTC (permalink / raw)
To: gcc-bugs
------- Comment #14 from ebotcazou at gcc dot gnu dot org 2006-10-31 17:56 -------
Subject: Bug 24071
Author: ebotcazou
Date: Tue Oct 31 17:55:32 2006
New Revision: 118262
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=118262
Log:
PR target/24071
* gthr-posix.h (__gthread_active_p): New implementation on Solaris.
* gthr-posix95.h (__gthread_active_p): Likewise.
Modified:
branches/gcc-4_2-branch/gcc/ChangeLog
branches/gcc-4_2-branch/gcc/gthr-posix.h
branches/gcc-4_2-branch/gcc/gthr-posix95.h
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24071
^ permalink raw reply [flat|nested] 10+ messages in thread
* [Bug target/24071] __gthread_active_p vs __gthread_once
[not found] <bug-24071-365@http.gcc.gnu.org/bugzilla/>
` (7 preceding siblings ...)
2006-10-31 17:56 ` ebotcazou at gcc dot gnu dot org
@ 2006-10-31 18:03 ` ebotcazou at gcc dot gnu dot org
2006-11-30 1:23 ` chaoyingfu at gcc dot gnu dot org
9 siblings, 0 replies; 10+ messages in thread
From: ebotcazou at gcc dot gnu dot org @ 2006-10-31 18:03 UTC (permalink / raw)
To: gcc-bugs
------- Comment #15 from ebotcazou at gcc dot gnu dot org 2006-10-31 18:02 -------
Hopefully put to rest on mainline and 4.2 branch.
--
ebotcazou at gcc dot gnu dot org changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|ASSIGNED |RESOLVED
Resolution| |FIXED
Target Milestone|--- |4.2.0
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24071
^ permalink raw reply [flat|nested] 10+ messages in thread
* [Bug target/24071] __gthread_active_p vs __gthread_once
[not found] <bug-24071-365@http.gcc.gnu.org/bugzilla/>
` (8 preceding siblings ...)
2006-10-31 18:03 ` ebotcazou at gcc dot gnu dot org
@ 2006-11-30 1:23 ` chaoyingfu at gcc dot gnu dot org
9 siblings, 0 replies; 10+ messages in thread
From: chaoyingfu at gcc dot gnu dot org @ 2006-11-30 1:23 UTC (permalink / raw)
To: gcc-bugs
------- Comment #16 from chaoyingfu at gcc dot gnu dot org 2006-11-30 01:23 -------
Subject: Bug 24071
Author: chaoyingfu
Date: Thu Nov 30 01:23:04 2006
New Revision: 119351
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=119351
Log:
Merged revisions 118230-118261 via svnmerge from
svn+ssh://chaoyingfu@sources.redhat.com/svn/gcc/trunk
........
r118230 | smw | 2006-10-31 09:04:32 -0800 (Tue, 31 Oct 2006) | 1 line
MAINTAINERS (Write After Approval): Add myself.
........
r118231 | charlet | 2006-10-31 09:43:17 -0800 (Tue, 31 Oct 2006) | 19 lines
2006-10-31 Robert Dewar <dewar@adacore.com>
* a-taster.adb, s-traent-vms.adb, a-elchha.ads, a-elchha.adb,
a-exctra.adb, ali-util.adb, exp_disp.ads, s-stalib.ads, s-traent.adb,
s-addope.ads, s-addope.adb, a-rbtgso.adb, a-crbltr.ads, a-coprnu.adb,
a-cgcaso.adb, a-cgarso.adb, a-cgaaso.adb, a-coormu.adb, a-ciormu.adb,
a-rbtgso.ads, a-stunha.adb, a-stunha.adb, a-ciorma.adb, a-coorma.adb,
a-secain.adb, a-slcain.adb, a-shcain.adb, a-stwiha.adb, a-stwiha.adb,
a-strhas.adb, a-strhas.adb, a-stzhas.adb, a-stzhas.adb, a-szuzha.adb,
a-chacon.adb, a-chacon.adb, a-chacon.ads, a-stboha.adb, a-swbwha.adb,
a-szbzha.adb: Minor reformatting. Fix header.
* a-numaux-x86.adb: Add parentheses for use of unary minus
* a-ngcefu.adb: Supply missing parentheses for unary minus
* a-ngcoty.adb: Add parens for use of unary minus
* a-ngelfu.adb: Add missing parens for unary minus
* a-tifiio.adb: Add parentheses for uses of unary minus
........
r118232 | charlet | 2006-10-31 09:44:22 -0800 (Tue, 31 Oct 2006) | 51 lines
2006-10-31 Robert Dewar <dewar@adacore.com>
Bob Duff <duff@adacore.com>
Ed Schonberg <schonberg@adacore.com>
* sem_res.adb (Resolve_Unary_Op): Add warning for use of unary minus
with multiplying operator.
(Expected_Type_Is_Any_Real): New function to determine from the Parent
pointer whether the context expects "any real type".
(Resolve_Arithmetic_Op): Do not give an error on calls to the
universal_fixed "*" and "/" operators when they are used in a context
that expects any real type. Also set the type of the node to
Universal_Real in this case, because downstream processing requires it
(mainly static expression evaluation).
Reword some continuation messages
Add some \\ sequences to continuation messages
(Resolve_Call): Refine infinite recursion case. The test has been
sharpened to eliminate some false positives.
Check for Current_Task usage now includes entry barrier, and is now a
warning, not an error.
(Resolve): If the call is ambiguous, indicate whether an interpretation
is an inherited operation.
(Check_Aggr): When resolving aggregates, skip associations with a box,
which are priori correct, and will be replaced by an actual default
expression in the course of expansion.
(Resolve_Type_Conversion): Add missing support for conversion from
a class-wide interface to a tagged type. Minor code cleanup.
(Valid_Tagged_Converion): Add support for abstact interface type
conversions.
(Resolve_Selected_Component): Call Generate_Reference here rather than
during analysis, and use May_Be_Lvalue to distinguish read/write.
(Valid_Array_Conversion): New procedure, abstracted from
Valid_Conversion, to incorporate accessibility checks for arrays of
anonymous access types.
(Valid_Conversion): For a conversion to a numeric type occurring in an
instance or inlined body, no need to check that the operand type is
numeric, since this has been checked during analysis of the template.
Remove legacy test for scope name Unchecked_Conversion.
* sem_res.ads: Minor reformatting
* a-except.adb, a-except-2005.adb: Turn off subprogram ordering
(PE_Current_Task_In_Entry_Body): New exception code
(SE_Restriction_Violation): Removed, not used
* a-except.ads: Update comments.
* types.h, types.ads: Add definition for Validity_Check
(PE_Current_Task_In_Entry_Body): New exception code
(SE_Restriction_Violation): Removed, not used
........
r118233 | charlet | 2006-10-31 09:44:38 -0800 (Tue, 31 Oct 2006) | 6 lines
2006-10-31 Thomas Quinot <quinot@adacore.com>
* g-socthi-vxworks.adb (C_Gethostbyname): Fix wrong test for returned
error status.
........
r118234 | charlet | 2006-10-31 09:44:55 -0800 (Tue, 31 Oct 2006) | 43 lines
2006-10-31 Hristian Kirtchev <kirtchev@adacore.com>
Jose Ruiz <ruiz@adacore.com>
* a-calend-vms.adb (Leap_Sec_Ops): Temp body for package in private
part of Ada.Calendar: all subprogram raise Unimplemented.
(Split_W_Offset): Temp function body, raising Unimplemented
* a-calend.ads, a-calend-vms.ads:
Add imported variable Invalid_TZ_Offset used to designate targets
unable
to support time zones.
(Unimplemented): Temporary function raised by the body of new
subprograms below.
(Leap_Sec_Ops): New package in the private part of Ada.Calendar. This
unit provides handling of leap seconds and is used by the new Ada 2005
packages Ada.Calendar.Arithmetic and Ada.Calendar.Formatting.
(Split_W_Offset): Identical spec to that of Ada.Calendar.Split. This
version returns an extra value which is the offset to UTC.
* a-calend.adb (Split_W_Offset): Add call to localtime_tzoff.
(Leap_Sec_Ops): New body for package in private part of Ada.Calendar.
(Split_W_Offset): New function body.
(Time_Of): When a date is close to UNIX epoch, compute the time for
that date plus one day (that amount is later substracted after
executing mktime) so there are no problems with time zone adjustments.
* a-calend-mingw.adb: Remove Windows specific version no longer needed.
* a-calari.ads, a-calari.adb, a-calfor.ads, a-calfor.adb,
a-catizo.ads, a-catizo.adb: New files.
* impunit.adb: Add new Ada 2005 entries
* sysdep.c: Add external variable __gnat_invalid_tz_offset.
Rename all occurences of "__gnat_localtime_r" to
"__gnat_localtime_tzoff".
(__gnat_localtime_tzoff for Windows): Add logic to retrieve the time
zone data and calculate the GMT offset.
(__gnat_localtime_tzoff for Darwin, Free BSD, Linux, Lynx and Tru64):
Use the field "tm_gmtoff" to extract the GMT offset.
(__gnat_localtime_tzoff for AIX, HPUX, SGI Irix and Sun Solaris): Use
the external variable "timezone" to calculate the GMT offset.
........
r118235 | charlet | 2006-10-31 09:45:11 -0800 (Tue, 31 Oct 2006) | 53 lines
2006-10-31 Arnaud Charlet <charlet@adacore.com>
Jose Ruiz <ruiz@adacore.com>
* s-osinte-posix.adb, s-osinte-linux.ads, s-osinte-freebsd.adb,
s-osinte-freebsd.ads, s-osinte-solaris-posix.ads, s-osinte-hpux.ads,
s-osinte-darwin.adb, s-osinte-darwin.ads, s-osinte-lynxos-3.ads,
s-osinte-lynxos-3.adb (To_Target_Priority): New function maps from
System.Any_Priority to a POSIX priority on the target.
* system-linux-ia64.ads:
Extend range of Priority types on Linux to use the whole range made
available by the system.
* s-osinte-aix.adb, s-osinte-aix.ads (To_Target_Priority): New
function maps from System.Any_Priority to a POSIX priority on the
target.
(PTHREAD_PRIO_PROTECT): Set real value.
(PTHREAD_PRIO_INHERIT): Now a function.
(SIGCPUFAIL): New signal.
(Reserved): Add SIGALRM1, SIGWAITING, SIGCPUFAIL, since these signals
are documented as reserved by the OS.
* system-aix.ads: Use the full range of priorities provided by the
system on AIX.
* s-taprop-posix.adb: Call new function To_Target_Priority.
(Set_Priority): Take into account Task_Dispatching_Policy and
Priority_Specific_Dispatching pragmas when determining if Round Robin
must be used for scheduling the task.
* system-linux-x86_64.ads, system-linux-x86.ads,
system-linux-ppc.ads: Extend range of Priority types on Linux to use
the whole range made available by the system.
* s-taprop-vms.adb, s-taprop-mingw.adb, s-taprop-irix.adb,
s-taprop-tru64.adb, s-taprop-linux.adb, s-taprop-hpux-dce.adb,
s-taprop-lynxos.adb (Finalize_TCB): invalidate the stack-check cache
when deallocating the TCB in order to avoid potential references to
deallocated data.
(Set_Priority): Take into account Task_Dispatching_Policy and
Priority_Specific_Dispatching pragmas when determining if Round Robin
or FIFO within priorities must be used for scheduling the task.
* s-taprop-vxworks.adb (Enter_Task): Store the user-level task id in
the Thread field (to be used internally by the run-time system) and the
kernel-level task id in the LWP field (to be used by the debugger).
(Create_Task): Reorganize to unify the calls to taskSpawn into a single
instance, and propagate the current task options to the spawned task.
(Set_Priority): Take into account Priority_Specific_Dispatching
pragmas.
(Initialize): Set Round Robin dispatching when the corresponding pragma
is in effect.
........
r118236 | charlet | 2006-10-31 09:45:31 -0800 (Tue, 31 Oct 2006) | 15 lines
2006-10-31 Robert Dewar <dewar@adacore.com>
* system-vms_64.ads, system-darwin-ppc.ads, system-vxworks-x86.ads,
system-linux-hppa.ads, system-hpux-ia64.ads,
system-lynxos-ppc.ads, system-lynxos-x86.ads, system-tru64.ads,
system-vxworks-sparcv9.ads, system-solaris-x86.ads,
system-irix-o32.ads, system-irix-n32.ads, system-hpux.ads,
system-vxworks-m68k.ads, system-vxworks-mips.ads, system-interix.ads,
system-solaris-sparc.ads, system-solaris-sparcv9.ads, system-vms.ads,
system-mingw.ads, system-vms-zcx.ads, system-vxworks-ppc.ads,
system-vxworks-alpha.ads, system.ads: Add pragma Warnings(Off,
Default_Bit_Order) to kill constant condition warnings for references
to this switch.
........
r118237 | charlet | 2006-10-31 09:45:59 -0800 (Tue, 31 Oct 2006) | 24 lines
2006-10-31 Vincent Celier <celier@adacore.com>
Eric Botcazou <ebotcazou@adacore.com>
* mlib-tgt-lynxos.adb, mlib-tgt-mingw.adb, mlib-tgt-tru64.adb,
mlib-tgt-aix.adb, mlib-tgt-irix.adb, mlib-tgt-hpux.adb,
mlib-tgt-linux.adb, mlib-tgt-solaris.adb: Use Append_To, instead of
Ext_To, when building the library file name
* mlib-tgt-vxworks.adb: ditto.
(Get_Target_Suffix): Add support for x86 targets.
* mlib-fil.ads, mlib-fil.adb: (Append_To): New function
* mlib-tgt-darwin.adb:
Use Append_To, instead of Ext_To, when building the library file name
(Flat_Namespace): New global variable.
(No_Shared_Libgcc_Switch): Rename to No_Shared_Libgcc_Options.
(Shared_Libgcc_Switch): Rename to With_Shared_Libgcc_Options.
(Link_Shared_Libgcc): Delete.
(Build_Dynamic_Library): Adjust for above changes.
Use Opt package.
(Build_Dynamic_Library): Pass -shared-libgcc if GCC 4 or later.
........
r118238 | charlet | 2006-10-31 09:46:26 -0800 (Tue, 31 Oct 2006) | 6 lines
2006-10-31 Eric Botcazou <ebotcazou@adacore.com>
* s-taprop-solaris.adb: (Time_Slice_Val): Change type to Integer.
(Initialize): Add type conversions required by above change.
........
r118239 | charlet | 2006-10-31 09:47:01 -0800 (Tue, 31 Oct 2006) | 13 lines
2006-10-31 Jose Ruiz <ruiz@adacore.com>
* s-osinte-vxworks.ads, s-osinte-vxworks.adb:
(getpid): New body for this function that uses the underlying
taskIdSelf
function for VxWorks 5 and VxWorks 6 in kernel mode.
(unsigned_int): New type, modular to allow logical bit operations.
(taskOptionsGet): New imported function.
* s-taspri-vxworks.ads (Private_Data): Change the type for the LWP
field to be compliant with the type used by the corresponding operating
system primitive.
........
r118240 | charlet | 2006-10-31 09:47:20 -0800 (Tue, 31 Oct 2006) | 30 lines
2006-10-31 Pascal Obry <obry@adacore.com>
Eric Botcazou <ebotcazou@adacore.com>
Vincent Celier <celier@adacore.com>
* adaint.c (__gnat_get_libraries_from_registry): Call explicitly the
ASCII version of the registry API. This is needed as the GNAT runtime
is now UNICODE by default.
Include version.h.
(get_gcc_version): Do not hardcode the return value.
(__gnat_file_time_name): On Windows properly set the default returned
value to -1 which corresponds to Invalid_Time.
(__gnat_fopen): New routine. A simple wrapper on all plateforms
except on Windows where it does conversion for unicode support.
(__gnat_freopen): Idem.
(__gnat_locate_exec_on_path): If environment variable PATH does not
exist, return a NULL pointer
* adaint.h: (__gnat_fopen): Declare.
(__gnat_freopen): Likewise.
* mingw32.h (_tfreopen): Define this macro here for older MingW
version.
Activate the unicode support on platforms using a MingW runtime
version 3.9 or newer.
* s-crtl.ads (fopen): Is now an import to the wrapper __gnat_freopen.
This is needed for proper unicode support on Windows.
(freopen): Idem.
........
r118241 | charlet | 2006-10-31 09:48:46 -0800 (Tue, 31 Oct 2006) | 15 lines
2006-10-31 Bob Duff <duff@adacore.com>
* a-filico.adb (Finalize(List_Controller)): Mark the finalization list
as finalization-started, so we can raise Program_Error on 'new'.
* s-finimp.adb: Raise Program_Error on 'new' if finalization of the
collection has already started.
* s-finimp.ads (Collection_Finalization_Started): Added new special
flag value for indicating that a collection's finalization has started.
* s-tassta.adb (Create_Task): Raise Program_Error on an attempt to
create a task whose master has already waited for dependent tasks.
........
r118242 | charlet | 2006-10-31 09:49:05 -0800 (Tue, 31 Oct 2006) | 12 lines
2006-10-31 Robert Dewar <dewar@adacore.com>
* lib.adb, lib.ads: (In_Predefined_Unit): New functions
* a-finali.ads, a-ngcoty.ads, a-strbou.ads, a-stream.ads, a-strmap.ads,
a-strunb.ads, a-stwibo.ads, a-stwima.ads, a-stwiun.ads, a-taside.ads,
a-coorse.ads, a-convec.ads, a-coinve.ads, a-cohama.ads, a-ciorse.ads,
a-cihama.ads, a-cihase.ads, a-cohase.ads, a-ciorma.ads, a-coorma.ads,
a-ciormu.ads, a-coormu.ads, a-stzbou.ads, a-stzmap.ads, a-stzunb.ads,
a-except-2005.ads: Add pragma Preelaborable_Warning
........
r118243 | charlet | 2006-10-31 09:49:53 -0800 (Tue, 31 Oct 2006) | 96 lines
2006-10-31 Robert Dewar <dewar@adacore.com>
Jose Ruiz <ruiz@adacore.com>
* a-dispat.ads, a-diroro.ads, a-diroro.adb: New files.
* ali.adb (Get_Name): Properly handle scanning of wide character names
encoded with brackets notation.
(Known_ALI_Lines): Add S lines to this list.
(Scan_ALI): Acquire S (priority specific dispatching) lines.
New flag Elaborate_All_Desirable in unit table
* ali.ads (Priority_Specific_Dispatching): Add this range of
identifiers to be used for Priority_Specific_Dispatching table entries.
(ALIs_Record): Add First_Specific_Dispatching and
Last_Specific_Dispatching that point to the first and last entries
respectively in the priority specific dispatching table for this unit.
(Specific_Dispatching): Add this table for storing each S (priority
specific dispatching) line encountered in the input ALI file.
New flag Elaborate_All_Desirable in unit table
* bcheck.adb: (Check_Configuration_Consistency): Add call to
Check_Consistent_Dispatching_Policy.
(Check_Consistent_Dispatching_Policy): Add this procedure in charge of
verifying that the use of Priority_Specific_Dispatching,
Task_Dispatching_Policy, and Locking_Policy is consistent across the
partition.
* bindgen.adb: (Public_Version_Warning): function removed.
(Set_PSD_Pragma_Table): Add this procedure in charge of getting the
required information from ALI files in order to initialize the table
containing the specific dispatching policy.
(Gen_Adainit_Ada): Generate the variables required for priority
specific
dispatching entries (__gl_priority_specific_dispatching and
__gl_num_specific_dispatching).
(Gen_Adainit_C): Generate the variables required for priority specific
dispatching entries (__gl_priority_specific_dispatching and
__gl_num_specific_dispatching).
(Gen_Output_File): Acquire settings for Priority_Specific_Dispatching
pragma entries.
(Gen_Restrictions_String_1, Gen_Restrictions_String_2): Removed.
(Gen_Restrictions_Ada, Gen_Restrictions_C, Set_Boolean): New
procedures.
(Tab_To): Removed.
(Gen_Output_File_Ada/_C): Set directly __gl_xxx variables instead of
a call to gnat_set_globals.
Generate a string containing settings from
Priority_Specific_Dispatching pragma entries.
(Gen_Object_Files_Options): Do not include the runtime libraries when
pragma No_Run_Time is specified.
* init.c (__gnat_install_handler, case FreeBSD): Use SA_SIGINFO, for
consistency with s-intman-posix.adb.
(__gnat_error_handler, case FreeBSD): Account for the fact that the
handler is installed with SA_SIGINFO.
(__gnat_adjust_context_for_raise, FreeBSD case): New function for
FreeBSD ZCX support, copied from Linux version.
Add MaRTE-specific definitions for the linux target. Redefine
sigaction,
sigfillset, and sigemptyset so the routines defined by MaRTE.
(__gl_priority_specific_dispatching): Add this variable that stores the
string containing priority specific dispatching policies in the
partition.
(__gl_num_specific_dispatching): Add this variable that indicates the
highest priority for which a priority specific dispatching pragma
applies.
(__gnat_get_specific_dispatching): Add this routine that returns the
priority specific dispatching policy, as set by a
Priority_Specific_Dispatching pragma appearing anywhere in the current
partition. The input argument is the priority number, and the result
is the upper case first character of the policy name.
(__gnat_set_globals): Now a dummy function.
(__gnat_handle_vms_condition): Feed adjust_context_for_raise with
mechargs instead of sigargs, as the latter can be retrieved from the
former and sigargs is not what we want on ia64.
(__gnat_adjust_context_for_raise, alpha-vms): Fetch sigargs from the
mechargs argument.
(__gnat_adjust_context_for_raise, ia64-vms): New function.
(tasking_error): Remove unused symbol.
(_abort_signal): Move this symbol to the IRIX specific part since this
is the only target that uses this definition.
(Check_Abort_Status): Move this symbol to the IRIX specific part since
this is the only target that uses this definition.
(Lock_Task): Remove unused symbol.
(Unlock_Task): Remove unused symbol.
* lib-writ.adb (Write_ALI): Output new S lines for
Priority_Specific_Dispatching pragmas.
Implement new flag BD for elaborate body desirable
* lib-writ.ads: Document S lines for Priority Specific Dispatching.
(Specific_Dispatching): Add this table for storing the entries
corresponding to Priority_Specific_Dispatching pragmas.
Document new BD flag for elaborate body desirable
* par-prag.adb (Prag): Add Priority_Specific_Dispatching to the list
of known pragmas.
........
r118244 | charlet | 2006-10-31 09:50:11 -0800 (Tue, 31 Oct 2006) | 47 lines
2006-10-31 Javier Miranda <miranda@adacore.com>
* a-tags.ads, a-tags.adb:
(Predefined_DT): New function that improves readability of the code.
(Get_Predefined_Prim_Op_Address, Set_Predefined_Prim_Op_Address,
Inherit_DT): Use the new function Predefined_DT to improve code
readability.
(Register_Interface_Tag): Update assertion.
(Set_Interface_Table): Update assertion.
(Interface_Ancestor_Tags): New subprogram required to implement AI-405:
determining progenitor interfaces in Tags.
(Inherit_CPP_DT): New subprogram.
* exp_disp.adb (Expand_Interface_Thunk): Suppress checks during the
analysis of the thunk code.
(Expand_Interface_Conversion): Handle run-time conversion of
access to class wide types.
(Expand_Dispatching_Call): When generating the profile for the
subprogram itype for a dispatching operation, properly terminate the
formal parameters chaind list (set the Next_Entity of the last formal
to Empty).
(Collect_All_Interfaces): Removed. This routine has been moved to
sem_util and renamed as Collect_All_Abstract_Interfaces.
(Set_All_DT_Position): Hidden entities associated with abstract
interface primitives are not taken into account in the check for
3.9.3(10); this check is done with the aliased entity.
(Make_DT, Set_All_DT_Position): Enable full ABI compatibility for
interfacing with CPP by default.
(Expand_Interface_Conversion): Add missing support for static
conversion
from an interface to a tagged type.
(Collect_All_Interfaces): Add new out formal containing the list of
abstract interface types to cleanup the subprogram Make_DT.
(Make_DT): Update the code to generate the table of interfaces in case
of abstract interface types.
(Is_Predefined_Dispatching_Alias): New function that returns true if
a primitive is not a predefined dispatching primitive but it is an
alias of a predefined dispatching primitive.
(Make_DT): If the ancestor of the type is a CPP_Class and we are
compiling under full ABI compatibility mode we avoid the generation of
calls to run-time services that fill the dispatch tables because under
this mode we currently inherit the dispatch tables in the IP
subprogram.
(Write_DT): Emit an "is null" indication for a null procedure
primitive.
(Expand_Interface_Conversion): Use an address as the type of the formal
of the internally built function that handles the case in which the
target type is an access type.
........
r118245 | charlet | 2006-10-31 09:50:31 -0800 (Tue, 31 Oct 2006) | 5 lines
2006-10-31 Robert Dewar <dewar@adacore.com>
* binde.adb (Better_Choice, Worse_Choice): Implement new preferences.
........
r118246 | charlet | 2006-10-31 09:50:47 -0800 (Tue, 31 Oct 2006) | 7 lines
2006-10-31 Robert Dewar <dewar@adacore.com>
* bindusg.ads, bindusg.adb:
Change to package and rename procedure as Display, which
now ensures that it only outputs usage information once.
........
r118247 | charlet | 2006-10-31 09:51:01 -0800 (Tue, 31 Oct 2006) | 9 lines
2006-10-31 Jose Ruiz <ruiz@adacore.com>
* cal.c: Use the header sys/time.h for VxWorks 6.2 or greater when
using RTPs.
* mkdir.c: Use a different version of mkdir for VxWorks 6.2 or greater
when using RTPs.
........
r118248 | charlet | 2006-10-31 09:51:20 -0800 (Tue, 31 Oct 2006) | 37 lines
2006-10-31 Robert Dewar <dewar@adacore.com>
Ed Schonberg <schonberg@adacore.com>
* treepr.adb: Use new subtype N_Membership_Test
* checks.ads, checks.adb: Add definition for Validity_Check
(Range_Or_Validity_Checks_Suppressed): New function
(Ensure_Valid): Test Validity_Check suppressed
(Insert_Valid_Check): Test Validity_Check suppressed
(Insert_Valid_Check): Preserve Do_Range_Check flag
(Validity_Check_Range): New procedure
(Expr_Known_Valid): Result of membership test is always valid
(Selected_Range_Checks): Range checks cannot be applied to
discriminants
by themselves. Disabling those checks must also be done for task types,
where discriminants may be used for the bounds of entry families.
(Apply_Address_Clause_Check): Remove side-effects if address expression
is non-static and is not the name of a declared constant.
(Null_Exclusion_Static_Checks): Extend to handle
Function_Specification.
Code cleanup and new error messages.
(Enable_Range_Check): Test for some cases of suppressed checks
(Generate_Index_Checks): Suppress index checks if index checks are
suppressed for array object or array type.
(Apply_Selected_Length_Checks): Give warning for compile-time detected
length check failure, even if checks are off.
(Ensure_Valid): Do not generate a check on an indexed component whose
prefix is a packed boolean array.
* checks.adb: (Alignment_Checks_Suppressed): New function
(Apply_Address_Clause_Check): New procedure, this is a completely
rewritten replacement for Apply_Alignment_Check
(Get_E_Length/Get_E_First_Or_Last): Add missing barrier to ensure that
we request a discriminal value only in case of discriminants.
(Apply_Discriminant_Check): For Ada_05, only call Get_Actual_Subtype
for
assignments where the target subtype is unconstrained and the target
object is a parameter or dereference (other aliased cases are known
to be unconstrained).
........
r118249 | charlet | 2006-10-31 09:51:38 -0800 (Tue, 31 Oct 2006) | 10 lines
2006-10-31 Robert Dewar <dewar@adacore.com>
* clean.adb, gnatname.adb, gnatsym.adb, prep.adb, prep.ads,
prepcomp.adb, prj.ads, prj-strt.adb, sem_maps.ads,
vms_conv.adb: Fix bad table increment values (much too small)
* table.adb (Realloc): Make sure we get at least some new elements
Defends against silly small values for table increment
........
r118250 | charlet | 2006-10-31 09:51:56 -0800 (Tue, 31 Oct 2006) | 33 lines
2006-10-31 Robert Dewar <dewar@adacore.com>
Ed Schonberg <schonberg@adacore.com>
Bob Duff <duff@adacore.com>
* einfo.ads, einfo.adb (Obsolescent_Warning): Now defined on all
entities. Move other fields around to make this possible
(Is_Derived_Type): Add missing call to Is_Type.
(Extra_Formals): New function for subprograms, entries, subprogram
types.
(Set_Extra_Formals): New procedure for subprograms, entries, subp
types.
(First_Formal_With_Extras): New function for subprogs, entries, subp
types.
(Write_Field28_Name): New procedure for node display of
"Extra_Formals".
Add node information for E_Return_Statement.
(Elaborate_Body_Desirable): New flag
(Is_Return_By_Reference_Type): Rename Is_Return_By_Reference_Type
to be Is_Inherently_Limited_Type, because return-by-reference has
no meaning in Ada 2005.
(E_Return_Statement): New entity kind.
(Return_Applies_To): Field of E_Return_Statement.
(Is_Return_Object): New flag in object entities.
(Is_Dynamic_Scope): Make it True for E_Return_Statement.
(Must_Have_Preelab_Init): New flag
(Known_To_Have_Preelab_Init): New flag
(Is_Formal_Object): Move from Sem_Ch8 body to Einfo
(Is_Visible_Formal): New flag on entities in formal packages.
(Low_Bound_Known): New flag
(Non_Limited_View, Set_Non_Limited_View): Add membership test agains
Incomplete_Kind.
(Write_Field17_Name): Correct spelling of Non_Limited_View. Add name
output when Id is an incomplete subtype.
........
r118251 | charlet | 2006-10-31 09:52:20 -0800 (Tue, 31 Oct 2006) | 58 lines
2006-10-31 Robert Dewar <dewar@adacore.com>
* errout.ads, errout.adb (Finalize): Implement switch -gnatd.m
Avoid abbreviation Creat
(Finalize): List all sources in extended mail source if -gnatl
switch is active.
Suppress copyright notice to file in -gnatl=f mode if -gnatd7 set
(Finalize): Implement new -gnatl=xxx switch to output listing to file
(Set_Specific_Warning_On): New procedure
(Set_Specific_Warning_Off): New procedure
Add implementation of new insertion \\
(Error_Msg_Internal): Add handling for Error_Msg_Line_Length
(Unwind_Internal_Type): Improve report on anonymous
access_to_subprogram
types.
(Error_Msg_Internal): Make sure that we set Last_Killed to
True when a message from another package is suppressed.
Implement insertion character ~ (insert string)
(First_Node): Minor adjustments to get better placement.
* frontend.adb:
Implement new -gnatl=xxx switch to output listing to file
* gnat1drv.adb:
Implement new -gnatl=xxx switch to output listing to file
* opt.ads: (Warn_On_Questionable_Missing_Paren): New switch
(Commands_To_Stdout): New flag
Implement new -gnatl=xxx switch to output listing to file
New switch Dump_Source_Text
(Warn_On_Deleted_Code): New warning flag for -gnatwt
Define Error_Msg_Line_Length
(Warn_On_Assumed_Low_Bound): New switch
* osint.ads, osint.adb
(Normalize_Directory_Name): Fix bug.
Implement new -gnatl=xxx switch to output listing to file
(Concat): Removed, replaced by real concatenation
Make use of concatenation now allowed in compiler
(Executable_Prefix.Get_Install_Dir): First get the full path, so that
we find the 'lib' or 'bin' directory even when the tool has been
invoked with a relative path.
(Executable_Name): New function taking string parameters.
* osint-c.ads, osint-c.adb:
Implement new -gnatl=xxx switch to output listing to file
* sinput-d.adb: Change name Creat_Debug_File to Create_Debug_File
* switch-c.adb:
Implement new -gnatl=xxx switch to output listing to file
Recognize new -gnatL switch
(no longer keep in old warning about old style usage)
Use concatenation to simplify code
Recognize -gnatjnn switch
(Scan_Front_End_Switches): Clean up handling of -gnatW
(Scan_Front_End_Switches): Include Warn_On_Assumed_Low_Bound for -gnatg
........
r118252 | charlet | 2006-10-31 09:53:20 -0800 (Tue, 31 Oct 2006) | 11 lines
2006-10-31 Robert Dewar <dewar@adacore.com>
* erroutc.ads, erroutc.adb (Set_Specific_Warning_On): New procedure
(Set_Specific_Warning_Off): New procedure
(Warning_Specifically_Suppressed): New function
(Validate_Specific_Warnings): New procedure
(Output_Msg_Text): Complete rewrite to support -gnatjnn
* err_vars.ads: Implement insertion character ~ (insert string)
........
r118253 | charlet | 2006-10-31 09:53:35 -0800 (Tue, 31 Oct 2006) | 28 lines
2006-10-31 Bob Duff <duff@adacore.com>
Ed Schonberg <schonberg@adacore.com>
* exp_aggr.adb (Build_Record_Aggr_Code): For extension aggregates, if
the parent part is a build-in-place function call, generate
assignments.
(Expand_Record_Aggregate): Call Convert_To_Assignments if any
components
are build-in-place function calls.
(Replace_Self_Reference): New subsidiary of
Make_OK_Assignment_Statement, to replace an access attribute that is a
self-reference into an access to the appropriate component of the
target object. Generalizes previous mechanism to handle self-references
nested at any level.
(Is_Self_Referential_Init): Remove, not needed.
(Is_Self_Referential_Init): New predicate to simplify handling of self
referential components in record aggregates.
(Has_Default_Init_Comps, Make_OK_Assignment_Statement): Add guard to
check for presence of entity before checking for self-reference.
(Has_Default_Init_Comps): Return True if a component association is a
self-reference to the enclosing type, which can only come from a
default initialization.
(Make_OK_Assignment_Statement): If the expression is of the form
Typ'Acc, where Acc is an access attribute, the expression comes from a
default initialized self-referential component.
(Build_Record_Aggr_Code): If the type of the aggregate is a tagged type
that has been derived from several abstract interfaces we must also
initialize the tags of the secondary dispatch tables.
........
r118254 | charlet | 2006-10-31 09:53:50 -0800 (Tue, 31 Oct 2006) | 34 lines
2006-10-31 Ed Schonberg <schonberg@adacore.com>
Thomas Quinot <quinot@adacore.com>
Javier Miranda <miranda@adacore.com>
Robert Dewar <dewar@adacore.com>
* exp_attr.adb:
(Expand_Access_To_Protected_Op): If the context indicates that an
access
to a local operation may be transfered outside of the object, create
an
access to the wrapper operation that must be used in an external
call.
(Expand_N_Attribute_Reference, case Attribute_Valid): For the AAMP
target, pass the Valid attribute applied to a floating-point prefix on
to the back end without expansion.
(Storage_Size): Use the new run-time function Storage_Size to retrieve
the allocated storage when it is specified by a per-object expression.
(Expand_N_Attribute_Reference): Add case for Attribute_Stub_Type.
Nothing to do here, the attribute has been rewritten during semantic
analysis.
(Expand_Attribute_Reference): Handle expansion of the new Priority
attribute
(Find_Fat_Info): Handle case of universal real
(Expand_Access_To_Protected_Op): Fix use of access to protected
subprogram from inside the body of a protected entry.
(Expand_Access_To_Protected_Op): Common procedure for the expansion of
'Access and 'Unrestricted_Access, to transform the attribute reference
into a fat pointer.
(Is_Constrained_Aliased_View): New predicate to help determine whether
a
subcomponent's enclosing variable is aliased with a constrained
subtype.
(Expand_N_Attribute_Reference, case Attribute_Constrained): For Ada_05,
test Is_Constrained_Aliased_View rather than Is_Aliased_View, because
an aliased prefix must be known to be constrained in order to use True
for the attribute value, and now it's possible for some aliased views
to be unconstrained.
........
r118255 | charlet | 2006-10-31 09:54:05 -0800 (Tue, 31 Oct 2006) | 7 lines
2006-10-31 Robert Dewar <dewar@adacore.com>
* exp_ch2.adb: Change Is_Lvalue to May_Be_Lvalue
(Expand_Entity_Reference): Correct error of not handling subprogram
formals in current_value processing.
........
r118256 | charlet | 2006-10-31 09:54:22 -0800 (Tue, 31 Oct 2006) | 63 lines
2006-10-31 Javier Miranda <miranda@adacore.com>
Robert Dewar <dewar@adacore.com>
Ed Schonberg <schonberg@adacore.com>
Gary Dismukes <dismukes@adacore.com>
* exp_ch3.ads, exp_ch3.adb (Expand_N_Object_Declaration): Do not
register in the final list objects containing class-wide interfaces;
otherwise we incorrectly register the tag of the interface in the final
list.
(Make_Controlling_Function_Wrappers): Add missing barrier to do not
generate the wrapper if the parent primitive is abstract. This is
required to report the correct error message.
(Expand_N_Subtype_Indication): Do validity checks on range
(Clean_Task_Names): If an initialization procedure includes a call to
initialize a task (sub)component, indicate that the procedure will use
the secondary stack.
(Build_Init_Procedure, Init_Secondary_Tags): Enable full ABI
compatibility for interfacing with CPP by default.
(Expand_N_Object_Declaration): Only build an Adjust call when the
object's type is a nonlimited controlled type.
* exp_ch3.adb: Add with and use of Exp_Ch6.
(Expand_N_Object_Declaration): Check for object initialization that is
a
call to build-in-place function and apply Make_Build_In_Place_Call_In_
Object_Declaration to the call.
(Freeze_Type): When the designated type of an RACW was not frozen at
the
point where the RACW was declared, validate the primitive operations
with respect to E.2.2(14) when it finally is frozen.
(Build_Initialization_Call,Expand_Record_Controller): Rename
Is_Return_By_Reference_Type to be Is_Inherently_Limited_Type, because
return-by-reference has no meaning in Ada 2005.
(Init_Secondary_Tags): Add missing call to Set_Offset_To_Top
to register tag of the immediate ancestor interfaces in the
run-time structure.
(Init_Secondary_Tags): Moved to the specification to allow the
initialization of extension aggregates with abstract interfaces.
(Build_Master_Renaming): Make public, for use by function declarations
whose return type is an anonymous access type.
(Freeze_Record_Type): Replace call to Insert_List_Before by call to
Insert_List_Before_And_Analyze after the generation of the specs
associated with null procedures.
(Expand_Tagged_Root): Update documentation in its specification.
(Init_Secondary_Tags): Update documentation.
(Build_Init_Procedure): If we are compiling under CPP full ABI compa-
tibility mode and the immediate ancestor is a CPP_Pragma tagged type
then generate code to inherit the contents of the dispatch table
directly from the ancestor.
(Expand_Record_Controller): Insert controller component after tags of
implemented interfaces.
(Freeze_Record_Type): Call new procedure Make_Null_Procedure_Specs to
create null procedure overridings when null procedures are inherited
from interfaces.
(Make_Null_Procedure_Specs): New procedure to generate null procedure
declarations for overriding null primitives inherited from interfaces.
(Is_Null_Interface_Procedure): New function in
Make_Null_Procedure_Specs.
(Make_Predefined_Primitive_Specs/Predefined_Primitive_Bodies): If the
immediate ancestor of a tagged type is an abstract interface type we
must generate the specification of the predefined primitives associated
with controlled types (because the dispatch table of the ancestor is
null and hence these entries cannot be inherited). This is required to
elaborate well the dispatch table.
........
r118257 | charlet | 2006-10-31 09:54:34 -0800 (Tue, 31 Oct 2006) | 30 lines
2006-10-31 Javier Miranda <miranda@adacore.com>
Ed Schonberg <schonberg@adacore.com>
Bob Duff <duff@adacore.com>
Gary Dismukes <dismukes@adacore.com>
Robert Dewar <dewar@adacore.com>
* exp_ch4.adb (Expand_N_Type_Conversion): Handle missing interface
type
conversion.
(Expand_N_In): Do validity checks on range
(Expand_Selected_Component): Use updated for of Denotes_Discriminant.
(Expand_N_Allocator): For "new T", if the object is constrained by
discriminant defaults, allocate the right amount of memory, rather than
the maximum for type T.
(Expand_Allocator_Expression): Suppress the call to Remove_Side_Effects
when the allocator is initialized by a build-in-place call, since the
allocator is already rewritten as a reference to the function result,
and this prevents an unwanted duplication of the function call.
Add with and use of Exp_Ch6.
(Expand_Allocator_Expresssion): Check for an allocator whose expression
is a call to build-in-place function and apply
Make_Build_In_Place_Call_In_Allocator to the call (for both tagged and
untagged designated types).
(Expand_N_Unchecked_Type_Conversion): Do not do integer literal
optimization if source or target is biased.
(Expand_N_Allocator): Add comments for case of an allocator within a
function that returns an anonymous access type designating tasks.
(Expand_N_Allocator): apply discriminant checks for access
discriminants of anonymous access types (AI-402, AI-416)
........
r118258 | charlet | 2006-10-31 09:54:50 -0800 (Tue, 31 Oct 2006) | 43 lines
2006-10-31 Bob Duff <duff@adacore.com>
Robert Dewar <dewar@adacore.com>
Gary Dismukes <dismukes@adacore.com>
Ed Schonberg <schonberg@adacore.com>
* exp_ch5.ads (Expand_N_Extended_Return_Statement): New procedure.
* exp_ch5.adb (Expand_N_Loop_Statement): Do validity checks on range
(Expand_N_Assignment_Statement): Call
Make_Build_In_Place_Call_In_Assignment if the right-hand side is a
build-in-place function call. Currently, this can happen only for
assignments that come from aggregates.
Add -gnatd.l --Use Ada 95 semantics for limited function returns,
in order to alleviate the upward compatibility introduced by AI-318.
(Expand_N_Extended_Return_Statement): Add support for handling the
return object as a build-in-place result.
(Expand_Non_Function_Return): Implement simple return statements nested
within an extended return.
(Enable_New_Return_Processing): Turn on the new processing of return
statements.
(Expand_Non_Function_Return): For a return within an extended return,
don't raise Program_Error, because Sem_Ch6 now gives a warning.
(Expand_N_Extended_Return_Statement): Implement AI-318
(Expand_Simple_Function_Return): Ditto.
(Expand_N_If_Statement): Handle new -gnatwt warning
(Expand_N_Case_Statement): Handle new -gnatwt warning
(Expand_N_Assignment): Handle assignment to the Priority attribute of
a protected object.
(Expand_N_Assignment_Statement): Implement -gnatVe/E to control
validity checking of assignments to elementary record components.
(Expand_N_Return_Statement): return Class Wide types on the secondary
stack independantly of their controlled status since with HIE runtimes,
class wide types are not potentially controlled anymore.
* expander.adb (Expand): Add case for new N_Extended_Return_Statement
node kind.
* exp_ch11.adb (Expand_N_Handled_Sequence_Of_Statements): Avoid
Expand_Cleanup_Actions in case of N_Extended_Return_Statement, because
it expects a block, procedure, or task. The return statement will get
turned into a block, and Expand_Cleanup_Actions will happen then.
........
r118259 | ebotcazou | 2006-10-31 09:54:58 -0800 (Tue, 31 Oct 2006) | 5 lines
PR target/24071
* gthr-posix.h (__gthread_active_p): New implementation on Solaris.
* gthr-posix95.h (__gthread_active_p): Likewise.
........
r118260 | charlet | 2006-10-31 09:55:05 -0800 (Tue, 31 Oct 2006) | 80 lines
2006-10-31 Robert Dewar <dewar@adacore.com>
Ed Schonberg <schonberg@adacore.com>
Bob Duff <duff@adacore.com>
Gary Dismukes <dismukes@adacore.com>
* exp_ch6.ads, exp_ch6.adb: Use new Validity_Check suppression
capability.
(Expand_Inlined_Call): Tagged types are by-reference types, and
therefore should be replaced by a renaming declaration in the expanded
body, as is done for limited types.
(Expand_Call): If this is a call to a function with dispatching access
result, propagate tag from context.
(Freeze_Subprogram): Enable full ABI compatibility for interfacing with
CPP by default.
(Make_Build_In_Place_Call_In_Assignment): New procedure to do
build-in-place when the right-hand side of an assignment is a
build-in-place function call.
(Make_Build_In_Place_Call_In_Allocator): Apply an unchecked conversion
of the explicit dereference of the allocator to the result subtype of
the build-in-place function. This is needed to satisfy type checking
in cases where the caller's return object is created by an allocator
for
a class-wide access type and the type named in the allocator is a
specific type.
(Make_Build_In_Place_Call_In_Object_Declaration): Apply an unchecked
conversion of the reference to the declared object to the result
subtype
of the build-in-place function. This is needed to satisfy type checking
in cases where the declared object has a class-wide type. Also, in the
class-wide case, change the type of the object entity to the specific
result subtype of the function, to avoid passing a class-wide object
without explicit initialization to the back end.
(Register_Interface_DT_Entry): Moved outside the body of
Freeze_Subprogram because this routine is now public; it is called from
Check_Dispatching_Overriding to handle late overriding of abstract
interface primitives.
(Add_Access_Actual_To_Build_In_Place_Call): New utility procedure for
adding an implicit access actual on a call to a build-in-place
function.
(Expand_Actuals): Test for an actual parameter that is a call to a
build-in-place function and apply
Make_Build_In_Place_Call_In_Anonymous_Context to the call.
(Is_Build_In_Place_Function): New function to determine whether an
entity is a function whose calls should be handled as build-in-place.
(Is_Build_In_Place_Function_Call): New function to determine whether an
expression is a function call that should handled as build-in-place.
(Make_Build_In_Place_Call_In_Allocator): New procedure for handling
calls to build-in-place functions as the initialization of an
allocator.
(Make_Build_In_Place_Call_In_Anonymous_Context): New procedure for
handling calls to build-in-place functions in contexts that do not
involve init of a separate object (for example, actuals of subprogram
calls).
(Make_Build_In_Place_Call_In_Object_Declaration): New procedure for
handling calls to build-in-place functions as the initialization of an
object declaration.
(Detect_Infinite_Recursion): Add explicit parameter Process to
instantiation of Traverse_Body to avoid unreferenced warning.
(Check_Overriding_Inherited_Interfaces): Removed.
(Register_Interface_DT_Entry): Code cleanup.
(Register_Predefined_DT_Entry): Code cleanup.
(Expand_Inlined_Call.Rewrite_Procedure_Call): Do not omit block around
inlined statements if within a transient scope.
(Expand_Inlined_Call.Process_Formals): When replacing occurrences of
formal parameters with occurrences of actuals in inlined body,
establish
visibility on the proper view of the actual's subtype for the body's
context.
(Freeze_Subprogram): Do nothing if we are compiling under full ABI
compatibility mode and we have an imported CPP subprogram because
for now we assume that imported CPP primitives correspond with
objects whose constructor is in the CPP side (and therefore we
don't need to generate code to register them in the dispatch table).
(Expand_Actuals): Introduce copy of actual, only if it might be a bit-
aligned selected component.
(Add_Call_By_Copy_Node): Add missing code to handle the case in which
the actual of an in-mode parameter is a type conversion.
(Expand_Actuals): If the call does not come from source and the actual
is potentially misaligned, let gigi handle it rather than rejecting the
(Expand_N_Subprogram_Body, Freeze_Subprogram): set subprograms
returning
Class Wide types as returning by reference independantly of their
controlled status since with HIE runtimes class wide types are not
potentially controlled anymore.
........
r118261 | charlet | 2006-10-31 09:55:21 -0800 (Tue, 31 Oct 2006) | 12 lines
2006-10-31 Ed Schonberg <schonberg@adacore.com>
* exp_ch9.adb (Update_Prival_Types): Simplify code for entity
references that are private components of the protected object.
(Build_Barrier_Function): Set flag Is_Entry_Barrier_Function
(Update_Prival_Subtypes): Add explicit Process argument to
Traverse_Proc
instantiation to deal with warnings.
(Initialize_Protection): If expression for priority is non-static, use
System_Priority as its expected type, in case the expression has not
been analyzed yet.
........
Removed:
branches/fixed-point/gcc/ada/a-calend-mingw.adb
Modified:
branches/fixed-point/ (props changed)
branches/fixed-point/ChangeLog
branches/fixed-point/MAINTAINERS
branches/fixed-point/gcc/ChangeLog
branches/fixed-point/gcc/ada/a-calend-vms.adb
branches/fixed-point/gcc/ada/a-calend-vms.ads
branches/fixed-point/gcc/ada/a-calend.adb
branches/fixed-point/gcc/ada/a-calend.ads
branches/fixed-point/gcc/ada/a-cgaaso.adb
branches/fixed-point/gcc/ada/a-cgarso.adb
branches/fixed-point/gcc/ada/a-cgcaso.adb
branches/fixed-point/gcc/ada/a-chacon.adb
branches/fixed-point/gcc/ada/a-chacon.ads
branches/fixed-point/gcc/ada/a-cihama.ads
branches/fixed-point/gcc/ada/a-cihase.ads
branches/fixed-point/gcc/ada/a-ciorma.adb
branches/fixed-point/gcc/ada/a-ciorma.ads
branches/fixed-point/gcc/ada/a-ciormu.adb
branches/fixed-point/gcc/ada/a-ciormu.ads
branches/fixed-point/gcc/ada/a-ciorse.ads
branches/fixed-point/gcc/ada/a-cohama.ads
branches/fixed-point/gcc/ada/a-cohase.ads
branches/fixed-point/gcc/ada/a-coinve.ads
branches/fixed-point/gcc/ada/a-convec.ads
branches/fixed-point/gcc/ada/a-coorma.adb
branches/fixed-point/gcc/ada/a-coorma.ads
branches/fixed-point/gcc/ada/a-coormu.adb
branches/fixed-point/gcc/ada/a-coormu.ads
branches/fixed-point/gcc/ada/a-coorse.ads
branches/fixed-point/gcc/ada/a-coprnu.adb
branches/fixed-point/gcc/ada/a-crbltr.ads
branches/fixed-point/gcc/ada/a-elchha.adb
branches/fixed-point/gcc/ada/a-elchha.ads
branches/fixed-point/gcc/ada/a-except-2005.adb
branches/fixed-point/gcc/ada/a-except-2005.ads
branches/fixed-point/gcc/ada/a-except.adb
branches/fixed-point/gcc/ada/a-except.ads
branches/fixed-point/gcc/ada/a-exctra.adb
branches/fixed-point/gcc/ada/a-filico.adb
branches/fixed-point/gcc/ada/a-finali.ads
branches/fixed-point/gcc/ada/a-ngcefu.adb
branches/fixed-point/gcc/ada/a-ngcoty.adb
branches/fixed-point/gcc/ada/a-ngcoty.ads
branches/fixed-point/gcc/ada/a-ngelfu.adb
branches/fixed-point/gcc/ada/a-numaux-x86.adb
branches/fixed-point/gcc/ada/a-rbtgso.adb
branches/fixed-point/gcc/ada/a-rbtgso.ads
branches/fixed-point/gcc/ada/a-secain.adb
branches/fixed-point/gcc/ada/a-shcain.adb
branches/fixed-point/gcc/ada/a-slcain.adb
branches/fixed-point/gcc/ada/a-stboha.adb
branches/fixed-point/gcc/ada/a-strbou.ads
branches/fixed-point/gcc/ada/a-stream.ads
branches/fixed-point/gcc/ada/a-strhas.adb
branches/fixed-point/gcc/ada/a-strmap.ads
branches/fixed-point/gcc/ada/a-strunb.ads
branches/fixed-point/gcc/ada/a-stunha.adb
branches/fixed-point/gcc/ada/a-stwibo.ads
branches/fixed-point/gcc/ada/a-stwiha.adb
branches/fixed-point/gcc/ada/a-stwima.ads
branches/fixed-point/gcc/ada/a-stwiun.ads
branches/fixed-point/gcc/ada/a-stzbou.ads
branches/fixed-point/gcc/ada/a-stzhas.adb
branches/fixed-point/gcc/ada/a-stzmap.ads
branches/fixed-point/gcc/ada/a-stzunb.ads
branches/fixed-point/gcc/ada/a-swbwha.adb
branches/fixed-point/gcc/ada/a-szbzha.adb
branches/fixed-point/gcc/ada/a-szuzha.adb
branches/fixed-point/gcc/ada/a-tags.adb
branches/fixed-point/gcc/ada/a-tags.ads
branches/fixed-point/gcc/ada/a-taside.ads
branches/fixed-point/gcc/ada/a-taster.adb
branches/fixed-point/gcc/ada/a-tifiio.adb
branches/fixed-point/gcc/ada/adaint.c
branches/fixed-point/gcc/ada/adaint.h
branches/fixed-point/gcc/ada/ali-util.adb
branches/fixed-point/gcc/ada/ali.adb
branches/fixed-point/gcc/ada/ali.ads
branches/fixed-point/gcc/ada/bcheck.adb
branches/fixed-point/gcc/ada/binde.adb
branches/fixed-point/gcc/ada/bindgen.adb
branches/fixed-point/gcc/ada/bindusg.adb
branches/fixed-point/gcc/ada/bindusg.ads
branches/fixed-point/gcc/ada/cal.c
branches/fixed-point/gcc/ada/checks.adb
branches/fixed-point/gcc/ada/checks.ads
branches/fixed-point/gcc/ada/clean.adb
branches/fixed-point/gcc/ada/einfo.adb
branches/fixed-point/gcc/ada/einfo.ads
branches/fixed-point/gcc/ada/err_vars.ads
branches/fixed-point/gcc/ada/errout.adb
branches/fixed-point/gcc/ada/errout.ads
branches/fixed-point/gcc/ada/erroutc.adb
branches/fixed-point/gcc/ada/erroutc.ads
branches/fixed-point/gcc/ada/exp_aggr.adb
branches/fixed-point/gcc/ada/exp_attr.adb
branches/fixed-point/gcc/ada/exp_ch11.adb
branches/fixed-point/gcc/ada/exp_ch2.adb
branches/fixed-point/gcc/ada/exp_ch3.adb
branches/fixed-point/gcc/ada/exp_ch3.ads
branches/fixed-point/gcc/ada/exp_ch4.adb
branches/fixed-point/gcc/ada/exp_ch5.adb
branches/fixed-point/gcc/ada/exp_ch5.ads
branches/fixed-point/gcc/ada/exp_ch6.adb
branches/fixed-point/gcc/ada/exp_ch6.ads
branches/fixed-point/gcc/ada/exp_ch9.adb
branches/fixed-point/gcc/ada/exp_disp.adb
branches/fixed-point/gcc/ada/exp_disp.ads
branches/fixed-point/gcc/ada/expander.adb
branches/fixed-point/gcc/ada/frontend.adb
branches/fixed-point/gcc/ada/g-socthi-vxworks.adb
branches/fixed-point/gcc/ada/gnat1drv.adb
branches/fixed-point/gcc/ada/gnatname.adb
branches/fixed-point/gcc/ada/gnatsym.adb
branches/fixed-point/gcc/ada/impunit.adb
branches/fixed-point/gcc/ada/init.c
branches/fixed-point/gcc/ada/lib-writ.adb
branches/fixed-point/gcc/ada/lib-writ.ads
branches/fixed-point/gcc/ada/lib.adb
branches/fixed-point/gcc/ada/lib.ads
branches/fixed-point/gcc/ada/mingw32.h
branches/fixed-point/gcc/ada/mkdir.c
branches/fixed-point/gcc/ada/mlib-fil.adb
branches/fixed-point/gcc/ada/mlib-fil.ads
branches/fixed-point/gcc/ada/mlib-tgt-aix.adb
branches/fixed-point/gcc/ada/mlib-tgt-darwin.adb
branches/fixed-point/gcc/ada/mlib-tgt-hpux.adb
branches/fixed-point/gcc/ada/mlib-tgt-irix.adb
branches/fixed-point/gcc/ada/mlib-tgt-linux.adb
branches/fixed-point/gcc/ada/mlib-tgt-lynxos.adb
branches/fixed-point/gcc/ada/mlib-tgt-mingw.adb
branches/fixed-point/gcc/ada/mlib-tgt-solaris.adb
branches/fixed-point/gcc/ada/mlib-tgt-tru64.adb
branches/fixed-point/gcc/ada/mlib-tgt-vxworks.adb
branches/fixed-point/gcc/ada/opt.ads
branches/fixed-point/gcc/ada/osint-c.adb
branches/fixed-point/gcc/ada/osint-c.ads
branches/fixed-point/gcc/ada/osint.adb
branches/fixed-point/gcc/ada/osint.ads
branches/fixed-point/gcc/ada/par-prag.adb
branches/fixed-point/gcc/ada/prep.adb
branches/fixed-point/gcc/ada/prep.ads
branches/fixed-point/gcc/ada/prepcomp.adb
branches/fixed-point/gcc/ada/prj-strt.adb
branches/fixed-point/gcc/ada/prj.ads
branches/fixed-point/gcc/ada/s-addope.adb
branches/fixed-point/gcc/ada/s-addope.ads
branches/fixed-point/gcc/ada/s-crtl.ads
branches/fixed-point/gcc/ada/s-finimp.adb
branches/fixed-point/gcc/ada/s-finimp.ads
branches/fixed-point/gcc/ada/s-osinte-aix.adb
branches/fixed-point/gcc/ada/s-osinte-aix.ads
branches/fixed-point/gcc/ada/s-osinte-darwin.adb
branches/fixed-point/gcc/ada/s-osinte-darwin.ads
branches/fixed-point/gcc/ada/s-osinte-freebsd.adb
branches/fixed-point/gcc/ada/s-osinte-freebsd.ads
branches/fixed-point/gcc/ada/s-osinte-hpux.ads
branches/fixed-point/gcc/ada/s-osinte-linux.ads
branches/fixed-point/gcc/ada/s-osinte-lynxos-3.adb
branches/fixed-point/gcc/ada/s-osinte-lynxos-3.ads
branches/fixed-point/gcc/ada/s-osinte-posix.adb
branches/fixed-point/gcc/ada/s-osinte-solaris-posix.ads
branches/fixed-point/gcc/ada/s-osinte-vxworks.adb
branches/fixed-point/gcc/ada/s-osinte-vxworks.ads
branches/fixed-point/gcc/ada/s-stalib.ads
branches/fixed-point/gcc/ada/s-taprop-hpux-dce.adb
branches/fixed-point/gcc/ada/s-taprop-irix.adb
branches/fixed-point/gcc/ada/s-taprop-linux.adb
branches/fixed-point/gcc/ada/s-taprop-lynxos.adb
branches/fixed-point/gcc/ada/s-taprop-mingw.adb
branches/fixed-point/gcc/ada/s-taprop-posix.adb
branches/fixed-point/gcc/ada/s-taprop-solaris.adb
branches/fixed-point/gcc/ada/s-taprop-tru64.adb
branches/fixed-point/gcc/ada/s-taprop-vms.adb
branches/fixed-point/gcc/ada/s-taprop-vxworks.adb
branches/fixed-point/gcc/ada/s-taspri-vxworks.ads
branches/fixed-point/gcc/ada/s-tassta.adb
branches/fixed-point/gcc/ada/s-traent-vms.adb
branches/fixed-point/gcc/ada/s-traent.adb
branches/fixed-point/gcc/ada/sem_maps.ads
branches/fixed-point/gcc/ada/sem_res.adb
branches/fixed-point/gcc/ada/sem_res.ads
branches/fixed-point/gcc/ada/switch-c.adb
branches/fixed-point/gcc/ada/sysdep.c
branches/fixed-point/gcc/ada/system-aix.ads
branches/fixed-point/gcc/ada/system-darwin-ppc.ads
branches/fixed-point/gcc/ada/system-hpux-ia64.ads
branches/fixed-point/gcc/ada/system-hpux.ads
branches/fixed-point/gcc/ada/system-interix.ads
branches/fixed-point/gcc/ada/system-irix-n32.ads
branches/fixed-point/gcc/ada/system-irix-o32.ads
branches/fixed-point/gcc/ada/system-linux-hppa.ads
branches/fixed-point/gcc/ada/system-linux-ia64.ads
branches/fixed-point/gcc/ada/system-linux-x86.ads
branches/fixed-point/gcc/ada/system-linux-x86_64.ads
branches/fixed-point/gcc/ada/system-lynxos-ppc.ads
branches/fixed-point/gcc/ada/system-lynxos-x86.ads
branches/fixed-point/gcc/ada/system-mingw.ads
branches/fixed-point/gcc/ada/system-solaris-sparc.ads
branches/fixed-point/gcc/ada/system-solaris-sparcv9.ads
branches/fixed-point/gcc/ada/system-solaris-x86.ads
branches/fixed-point/gcc/ada/system-tru64.ads
branches/fixed-point/gcc/ada/system-vms-zcx.ads
branches/fixed-point/gcc/ada/system-vms.ads
branches/fixed-point/gcc/ada/system-vms_64.ads
branches/fixed-point/gcc/ada/system-vxworks-alpha.ads
branches/fixed-point/gcc/ada/system-vxworks-m68k.ads
branches/fixed-point/gcc/ada/system-vxworks-mips.ads
branches/fixed-point/gcc/ada/system-vxworks-ppc.ads
branches/fixed-point/gcc/ada/system-vxworks-sparcv9.ads
branches/fixed-point/gcc/ada/system-vxworks-x86.ads
branches/fixed-point/gcc/ada/system.ads
branches/fixed-point/gcc/ada/table.adb
branches/fixed-point/gcc/ada/treepr.adb
branches/fixed-point/gcc/ada/types.ads
branches/fixed-point/gcc/ada/types.h
branches/fixed-point/gcc/ada/vms_conv.adb
branches/fixed-point/gcc/gthr-posix.h
branches/fixed-point/gcc/gthr-posix95.h
Propchange: branches/fixed-point/
('svnmerge-integrated' modified)
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24071
^ permalink raw reply [flat|nested] 10+ messages in thread
end of thread, other threads:[~2006-11-30 1:23 UTC | newest]
Thread overview: 10+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
[not found] <bug-24071-365@http.gcc.gnu.org/bugzilla/>
2005-10-02 21:21 ` [Bug libstdc++/24071] solaris vs. __gthread_active_p mark at codesourcery dot com
2005-10-02 21:28 ` ebotcazou at gcc dot gnu dot org
2005-10-09 17:57 ` [Bug target/24071] __gthread_active_p vs __gthread_once ebotcazou at gcc dot gnu dot org
2005-12-14 11:47 ` neil at fnxweb dot com
2006-10-28 5:36 ` ebotcazou at gcc dot gnu dot org
2006-10-31 9:34 ` bkoz at gcc dot gnu dot org
2006-10-31 17:55 ` ebotcazou at gcc dot gnu dot org
2006-10-31 17:56 ` ebotcazou at gcc dot gnu dot org
2006-10-31 18:03 ` ebotcazou at gcc dot gnu dot org
2006-11-30 1:23 ` chaoyingfu at gcc dot gnu dot org
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).